Get the app
← back to catalog · registered 2026-08-22 13:56

DreamFoundries/SuperGemma-4-12b-abliterated-8bit

DreamFoundries Gemma 12B second-order
Your rig guess connected
? Why do I need an app?
Reading your rig…

This is a rough estimate. Install the free app - we'll show exact numbers.

Reading real hardware from your app right now. Numbers below are exact.

Below is the per-quantization compatibility for this model. 

curl -H "Authorization: Bearer $ABL_KEY" \
     "https://abliteration.org/api/v1/models/DreamFoundries%2FSuperGemma-4-12b-abliterated-8bit"
Response includes
  • classification m1
  • files 12
  • hub_downloads_all_time 486
  • author_summary 12 models
  • readme_text full
Try next
GET/models?author=DreamFoundries
All 12 models by DreamFoundries
 GET/models?family=gemma&method=m1
Sibling gemma models · same method
 GET/methods/m1
What "m1" means · taxonomy
10 credits · hourly refresh · ~4 KB payload Get an API key →
Abliteration classifier · v1.0.0
M1
Primary method

Direct removal

No other method signals detected in this model.
Confidence
MEDIUM
Why this label 3 signals
Method inferred from partial signals - repository name, related files, or tag patterns. Producer identity not confirmed; label may sharpen or shift as we gather more evidence.
  • 'abliterated' in name/tags
  • is_gguf=0 (base model)
  • no specific method indicators - defaulting to M1 (most common)
How we label → About M1 →
Refusal direction extracted via
Extraction technique

Difference-of-means

FromArditi et al. (2024) ↗
Confidence
MEDIUM
Why we say so
primary_method=M1; difference-of-means is the reference extraction for M1/M3 (Arditi 2024)
What is a refusal direction? → Read the paper ↗
Downloads · lifetime
486
135 last 30d - stable
Likes
0
Model age
3mo ago
created 2026-06-29
Downloads over time
Now526→from122↑331%
102257412566122 on Jul 1526 on Oct 11526 on Oct 10JulAugSepOct
Jul 1 → Oct 11 · 54 snapshots · spans 102 days

Genealogy 0 direct forks

Full fork graph →

This model's place in the market. Above: what it was derived from. Below: the tree of everything derived from it.

Jiunsong/SuperGemma-4-12b-abliterated↓ 74 this lineage
DreamFoundries/SuperGemma-4-12b-abliterated-8bit↓ 135

Metadata

License
apache-2.0
Tags
mlx safetensors gemma4 mlx-lm quantized 8-bit base_model:Jiunsong/SuperGemma-4-12b-abliterated text-generation conversational base_model:quantized:Jiunsong/SuperGemma-4-12b-abliterated license:apache-2.0 region:us

Related

Other by DreamFoundries
SuperGemma-4-12b-abliterated-4bit 250 supergemma4-e4b-abliterated-6bit 225 supergemma4-e4b-abliterated-4bit 194 supergemma4-26b-abliterated-multimodal-8bit 190 SuperGemma-4-12b-abliterated-6bit 181
Same base model
mradermacher/SuperGemma-4-12b-abliterated-i1-GGUF 2K mradermacher/SuperGemma-4-12b-abliterated-GGUF 675 Jiunsong/SuperGemma-4-12b-abliterated-gguf-4bit 488 DreamFoundries/SuperGemma-4-12b-abliterated-4bit 250 Jiunsong/SuperGemma-4-12b-abliterated-mlx-4bit 191
Total size
11.8 GB
Files
12
Quantizations
1
Registered
2026-08-22 13:56
Last updated on HF
2026-08-25 19:19
Source huggingface.co/DreamFoundries/SuperGemma-4-12b-abliterated-8bit ↗ Base model Jiunsong/SuperGemma-4-12b-abliterated →

Files by quantization

Auxiliary files 12 files 11.8 GB
model-00002-of-00003.safetensors 4.99 GB 1bee1874 download
model-00001-of-00003.safetensors 4.98 GB 45284089 download
model-00003-of-00003.safetensors 1.81 GB 6b8cd7bc download
tokenizer.json 30.7 MB cc8d3a0c download
model.safetensors.index.json 131 KB 361a9740 download
chat_template.original.jinja 28.3 KB eceb4aaa download
config.json 4.47 KB 84f0912b download
generation_config.json 4.01 KB f28cf89e download
tokenizer_config.json 2.84 KB df65a194 download
README.md 2.18 KB 43ea2ac7 download
.gitattributes 1.53 KB 52373fe2 download
chat_template.jinja 522 B cede33d8 download

README current version from Hugging Face

library_name: mlx
base_model: Jiunsong/SuperGemma-4-12b-abliterated
license: apache-2.0
pipeline_tag: text-generation
tags:

  • mlx
  • mlx-lm
  • quantized
  • 8-bit
  • base_model:Jiunsong/SuperGemma-4-12b-abliterated

SuperGemma-4-12b-abliterated MLX 8-bit

This repository contains an MLX-LM conversion of Jiunsong/SuperGemma-4-12b-abliterated.

Conversion Details

  • Original model: Jiunsong/SuperGemma-4-12b-abliterated
  • Model family: SuperGemma4
  • Source model type: gemma4_unified
  • Model size: 11,959,730,224 parameters
  • Quantization: MLX-LM affine quantization
  • Bits: 8-bit
  • Group size: 64
  • Local MLX folder size at upload time: 11.81 GiB
  • Local safetensors weight size at upload time: 11.78 GiB

This Gemma conversion follows the MLX-LM Gemma 4 shared-KV topology and uses non-strict checkpoint loading so extra HF tensors outside that topology are discarded during conversion.

The temporary conversion config remapped gemma4_unified to gemma4 because this mlx-lm build exposes that Gemma 4 implementation.

For mlx-swift compatibility, per_layer_model_projection was left unquantized while the rest of the eligible linear layers were quantized.

For oMLX / mlx-swift chat-template compatibility, the published MLX package uses a simplified Gemma 4 chat template based on the <|turn>role\n...<turn|> format. The original upstream chat_template.jinja is preserved as chat_template.original.jinja.

Abliteration credit: the abliterated source model was published by Jiunsong. This repository only provides the MLX conversion and quantization.

Usage

mlx_lm.generate --model DreamFoundries/SuperGemma-4-12b-abliterated-8bit --prompt "Hello" --max-tokens 64

Benchmarks

No comparative benchmarks have been run yet. The repository does not currently provide quality, speed, memory, or benchmark comparisons against the original weights or other quantizations.

License

This is a converted/quantized derivative of the original model. Please refer to the original model repository for the upstream license and usage terms: https://huggingface.co/Jiunsong/SuperGemma-4-12b-abliterated

README history 4 versions

The author's README evolved over time. Click a version to see its content at that point.

  1. 2026-08-25Mueve Download MLXHub al final del documentoba900dc2.5 KB
    Loading...
  2. 2026-08-25Add Open in MLXHub / Download MLXHub badgesa201de22.5 KB
    Loading...
  3. 2026-08-25Añade badges Open in MLXHub / Download MLXHub6304cbc2.6 KB
    Loading...
  4. 2026-06-29Upload README.md3d9200d2.2 KB
    Loading...
Catalog is the map. Apps are the tools.

Run models on your own machine, not in the cloud.

Every model page has an "Open in Abliteration" button that hands the model directly to the first-party desktop client, at the quantization your rig can actually run. No API keys, no subscription, no prompt leakage.

Get the app Browse the catalog
Open in Abliteration